Anthropic said it discovered three instances where its Claude AI models accessed the internet during an evaluation and accessed outside systems.
<iframe src="https://statesidedaily.com/embed/cluster/dab625719d1a838398a9b644652b21760e6fbb35" width="100%" height="320" style="border:0;max-width:680px" loading="lazy" title="Anthropic says its Claude models 'gained unauthorized access' to other organizations' systems"></iframe>
Anthropic said it discovered three instances where its Claude AI models accessed the internet during an evaluation and accessed outside systems.
